TOWARDS AN INTELLIGENT TUTOR SYSTEM BASED ON ADAPTIVE WORKFLOWS

In this paper, we begin the design of an Intelligent Tutor System (ITS) based on adaptive workflows. The main idea of this work is to integrate a pedagogical activity base and an adaptive workflow engine into the general architecture of the ITS. This allows the generation, execution and follow-up of the completion of intelligent workflow models used for the representation of pedagogical sequences and self-learning processes.

The model of adapting pedagogical sequences and content to the preferences and learning level of the student we propose is based on the intelligent production of processes and their adaptations as they are executed based on certain parameters of the learner model and the domain of teaching schema adopted.

To achieve our objective, a meta-model of the process has been designed to generate adaptive models of pedagogical sequences. Thus the workflow engine allows to generate a model of the pedagogical sequence corresponding to the defined meta-model.

The adaptation of the sequence when it is executed is based on the learner interactions with the system collected by a supervisor agent.
The system thus defined allows the coexistence of an intelligent tutor to carry out the tasks of teachers in the teaching process and a human tutor who takes part in the performance of certain pedagogical tasks as well as the monitoring and follow-up of the execution of the models of self-learning processes and the resolution of the blocking and demotivating situations of the learners.